AGGREGATE ACREAGE RESPONSE OF CASHEW NUT AND SESAME TO COMMODITY PRICE AND NON PRICE FACTORS IN SOUTHEASTERN TANZANIA

Transcription

1 AGGREGATE ACREAGE RESPONSE OF CASHEW NUT AND SESAME TO COMMODITY PRICE AND NON PRICE FACTORS IN SOUTHEASTERN TANZANIA DEVIS FABIAN A DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN AGRICULTURAL ECONOMICS OF SOKOINE UNIVERSITY OF AGRICULTURE. MOROGORO, TANZANIA. 2013

2 ii ABSTRACT This study aimed at investigating the impact of price and non-price factors on cashew and sesame acreage in Nachingwea and Mtwara rural Districts. Growth rates analyses were also conducted through linearlized exponential growth model to trace the trends for area, yield (productivity) and production of the two crops for the period The general trend showed positive growth rates in area, yield and production, but with few exceptions. Meanwhile, the logarithmic functional form of the linear Nerlovian adjustment model was employed on time series data from to estimate acreage response to price and non-price factors. Results revealed that there was a positive and significant relationship of sesame acreage with price, and non-significant relationship with rainfall (non-price factor) in both districts. Similarly, cashew acreage was observed to have positive and significant relationship with price and positive but insignificant relationship with rainfall in Nachingwea District. In Mtwara rural, positive relationship existed between cashew acreage and price, while with rainfall the relationship was negative. The study further established that short and long run price elasticities of sesame acreage were and 0.515, respectively in Nachingwea, whereas short and long run price elasticities in Mtwara rural were and 1.65, respectively, which implies that farmers are more responsive to price changes in the long run than they are in the short run. Similarly, non-price short and long run elasticities for sesame were and 0.06, respectively in Nachingwea; and and 0.11, respectively for Mtwara rural District. For cashew acreage, the short and long run price elasticities were and 1.364, respectively for Nachingwea while in Mtwara rural short and long run price elasticities were 0.37 and 0.885, respectively. Meanwhile the study found that short and long run non price elasticities for Nachingwea were and 0.049, respectively. However, the

3 iii elasticities though positive, they generally fell in the inelastic range suggesting that price incentive in itself is essential but not sufficient.

4 iv DECLARATION I, DEVIS FABIAN, do hereby declare to the Senate of Sokoine University of Agriculture, that this dissertation is my own work done within the period of registration and that it has neither been submitted nor being concurrently submitted for degree award to any other institution. Devis Fabian Date (MSc. Candidate) The above declaration is confirmed Prof. A. C. Isinika Date (Supervisor)

5 v COPYRIGHT All rights are reserved. No part of this dissertation may be reproduced, stored in any retrieval system, or transmitted in any form or by any means without prior written permission of the author or Sokoine University of Agriculture in that behalf.

6 vi ACKNOWLEDGEMENTS In the first place, I would like to thank God the Almighty who gave me life and grace to venture into this undertaking. To Him, all praise is due. This work has been accomplished under the inspiring supervision of Prof. A.C. Isinika. She has been a constant source of inspiration from the time of topic selection and proposal development, all the way to the production of this dissertation. I would like to acknowledge her valuable comments and useful recommendations without which this study wouldn t have reached this far. I would also like to thank my sponsor, the government of Tanzania through the Commission for Science and Technology (COSTECH) for giving me the opportunity to climb another rung of the academic ladder. The provision of funds by the commission is highly appreciated. My acknowledgement also extends to Dr. Elly Kafiriti, Zonal Director for research and development (Southern zone), Dr. G.Mkamilo, Bernadetha Kimata and other staff of Naliendele Agricultural Research Institute who have, in various ways, contributed to the success of this undertaking. Deep appreciation also goes to extension personnel of Nachingwea and Mtwara rural District councils who, all heartedly, agreed to work with me during data collection. Without their willingness, the data for this work could not have been accessed. My beloved wife Sarah together with our children Dorcas and Hans have also remained a source of motivation for my academic achievements. They, most of the time, had to do without my presence but remained supportive to me. To them I say thank you.

7 vii DEDICATION This work is dedicated to my guardians, Mr. Gwalisu Gwaselya and Mrs. Esther Gwaselya who sowed the seed of education in me and made sure it grew to fruition.

8 viii TABLE OF CONTENTS ABSTRACT... ii DECLARATION... iv COPYRIGHT... v ACKNOWLEDGEMENTS... vi DEDICATION... vii TABLE OF CONTENTS... viii LIST OF TABLES... xi LIST OF FIGURES... xii LIST OF APPENDICES... xiii LIST OF ABBREVIATIONS AND ACRONYMS... xiv CHAPTER ONE INTRODUCTION Background Information Problem Statement and Justification Objectives of the Study General objective Specific objectives: Hypotheses... 6 CHAPTER TWO LITERATURE REVIEW Agriculture Sector in Tanzania Agricultural Policy in Tanzania... 9

9 ix 2.3 Sesame Production Cashew nut Sub-sector in Tanzania Growth Rate Analysis and Acreage Response Compound growth rates in agriculture Acreage response and estimation techniques Variables used in acreage response studies Dependent variables Independent variables CHAPTER THREE METHODOLOGY Conceptual Framework Research Design Choice of Crop and the Study Area Description of the Study Area Area and location Demography Climate and topography Economic activities Data Type and Sources Data Analysis Growth rate analysis Nerlovian model specification and estimation method... 31

10 x CHAPTER FOUR RESULTS AND DISCUSSION Growth Rates of Area, Yield and Production Test for Stationarity Acreage Response of Sesame Adjustment and expectation coefficients for sesame Short run and long run elasticities for sesame Acreage Response of Cashew Nut Adjustment and expectation coefficients for cashew Short run and long run elasticities for cashew Summary of Main Findings Growth rates Acreage response to price and non price factors CHAPTER FIVE CONCLUSIONS AND RECOMMENDATIONS Conclusions Recommendations REFERENCES APPENDICES... 63

11 xi LIST OF TABLES Table 1: Major export crops of Tanzania in Table 2: World s major sesame producers Table 3: Growth Rates of Area, Production and Yield of Sesame and cashew Table 4: Acreage response of sesame to price and non-price factors Table 5: Adjustment coefficient, short run and long run elasticities for sesame Table 6: Acreage response of cashew nut to price and non-price factors Table 7: Adjustment coefficients, short and long run elasticities for cashew nut... 47

12 xii LIST OF FIGURES Figure 1: Trend of exchange rates of the Tanzanian shillings against the USD Figure 2: Production of cashew nut in Tanzania: Figure 3: A map of Southeastern Tanzania showing sample districts (Nachingwea and Mtwara rural) Figure 4: Sesame Acreage, Production and Price in Nachingwea Figure 5: Sesame acreage, Production and Price in Mtwara rural Figure 6: Cashew acreage, Production and Price in Nachingwea Figure 7: Cashew acreage, Production and Price in Mtwara rural... 40

13 xiii LIST OF APPENDICES Appendix 1: Augmented Dickey-Fuller test statistics of unit roots for Nachingwea Appendix 2: Augmented Dickey-Fuller test statistics of unit roots for Mtwara rural Appendix 3: Area under sesame production by districts Appendix 4: Volume of sesame production by districts Appendix 5: Sesame market prices by districts Appendix 6: Area under cashew nut production by districts Appendix 7: Volume of cashew nut production by districts Appendix 8: Cashew nut market prices by districts... 68

14 xiv LIST OF ABBREVIATIONS AN ACRONYMS ADF ASDP CAADP CATA CBT COSTECH DADPs DALDO FAO GDP IMF LGAs Augmented Dickey-Fuller Agricultural Sector Development Programme Comprehensive Africa Agriculture Development Programme Cashew nut Authority of Tanzania Cashew nut Board of Tanzania Commission for Science and Technology District Agricultural Development Programmes District Agricultural and Livestock Officer Food and Agriculture Organization Gross Domestic Product International Monetary Fund Local Government Authorities MKUKUTA Mkakati wa Kukuza Uchumi na Kupunguza Umaskini Tanzania NARI PAFC SAGCOT TCMB URT Naliendele Agricultural Research Institute Partial Autocorrelation Function Southern Agricultural Growth Corridor of Tanzania Tanzania Cashew nut Marketing Board United Republic of Tanzania

15 1 CHAPTER ONE 1.0 INTRODUCTION 1.1 Background Information The contribution of agriculture sector to the economy in Tanzania remains relatively high, accounting for about 23.7% of the Gross Domestic Product (GDP) (URT, 2012), and employing majority of the rural population which is generally characterized by smallholders (URT, 2009). Hence, its performance in terms of productivity and total output has a significant effect on food security, income and livelihoods. According to URT (2009), the sector accounts for about 50 % of rural household incomes. Despite growth in other sectors such as mining and tourism, their contribution to poverty reduction remains low because they have weak backward and forward linkages to rural sources of livelihoods (Ibid). Hence, any plan to reduce poverty and improve the economy will necessarily have to include strategies to improve agriculture. Due to the importance of agriculture to the economy, in 2005 the government of Tanzania adapted the Agricultural Sector Development Programme (ASDP), which has two main objectives; (i) To enable farmers to have better access to, and foster improvement of agricultural knowledge, technologies, marketing systems and infrastructure all contributing to higher productivity, profitability and farm income, and (ii) to promote private investments in agriculture based on an improved regulatory and policy environment (URT, 2005). Effective implementation of the ASDP requires that farmers and extension staff who support them develop farm projects and enterprises that improve productivity such that farmers get positive returns from their investment, and they use such returns to improve

16 2 their families livelihoods. For example, at the district level, District Agricultural Development Programmes (DADPs) should be formulated based on technical and economic information regarding the profitability of different enterprises that are recommended to farmers for adoption. Statistics on acreage responsiveness to prevailing policy and other structural changes serve as useful guides when formulating agricultural policies, programmes and projects. If, for instance, production of a particular crop is responsive to price incentives, then policies which lead to increased commodity prices are expected to motivate farmers to allocate more resources for agricultural production. If the markets operate efficiently and farmers are assured of favourable prices, then they willingly adopt improved agronomic practices. Production of food and cash crops will increase and thus meet not only food requirements of the country, but it could also improve foreign exchange earnings through exports. Therefore, any agricultural price policy should aim at achieving fair income levels for farmers and increasing exports. However, Mumbengegwi (1990) and Muchapondwa (2008) argued that the use of agricultural policy instruments to affect agricultural activities at the farm level and performance, without empirical knowledge of the structural parameters of commodity supply response leaves the possibility that the policy instruments may be inappropriately used and thus affecting the agricultural sector negatively. Also, there is an argument that the area of land demanded by a farmer is a reflection of the expected economic incentives (Bridges and Tenkorang, 2009). For instance, since farmland is shared among different crops, changes in relative anticipated economic incentives leads to changes in the acreage demanded for each crop. Since land is the primary resource that is used for agricultural production, there was a need to know how farmers respond to policy changes in terms of acreage adjustment

17 3 when faced with different conditions as dictated by policy, programme and project changes, that are prescribed by different levels of government (viz local, district, national and international levels). This study intended to provide information regarding farmers acreage response as they face changing commodity prices of cashew and sesame in Southeastern Tanzania. Farmers response to economic incentives determines agriculture s contribution to the economy especially where the sector is the largest employer (Rahji et al., 2008). Such information is important for policy makers so that they make informed decisions in relation to crop production, marketing and related inputs during policy processes. 1.2 Problem Statement and Justification As stated earlier, one of the objectives of ASDP was to contribute to agricultural productivity. Traditionally, the national economy has been heavily reliant on agricultural growth and export earnings. In order for the national economy to grow, agricultural activities should be stimulated to increase farmers purchasing power thereby increasing domestic market for non-agricultural products to rural communities and rising foreign exchange earnings through agricultural exports. The contribution of the agricultural sector in these areas will depend on the responsiveness of agricultural production to various incentives including producer prices. For that reason, any pricing policy that is formulated in favour of the agricultural sector would require empirical knowledge on the supply response of the agricultural production. However, only a few studies have been conducted to assess the responsiveness of crop production to various factors, and these have been conducted at national level. For instance, McKay et al. (2006) addressed aggregate export and food crop supply response to liberalization of agricultural prices and marketing in Tanzania. Another study

18 4 by Eriksson (1993) addressed peasant response to price incentives in Tanzania. These studies concluded that where liberalization of agricultural markets increases the effective prices paid to farmers can be useful in promoting farm level improvements in production. There is no study that has focused on acreage response for crops to commodity price and non-price factor changes in Southeast Tanzania. This means, decisions on resource allocation in Lindi and Mtwara Regions have been based purely on experience of the planning officers, some of whom are fairly new to their work environment. Meanwhile, Tanzania has many graduates at the district level who can be used to improve the formulation and planning of DADPs using research based information which provides guidance on what response to expect from farmers, given certain policy or structural changes. Cashew nut and sesame constitute two major cash crops which play an important role as income sources for majority of farmers in the zone. According to the national sample census of agriculture conducted in 2007/08, Mtwara Region was the leading cashew producer, contributing to 35% of total planted area followed by Lindi (23%) (URT, 2012). Likewise, sesame is known to be the main export oriented oilseed crop in Lindi and Mtwara Regions, accounting for 35% of the total sesame export in Tanzania (Schul and Mburi, 2010). Enhanced cashew and sesame producers income would, therefore, be expected to encourage investment and adoption of improved techniques for the two crops. This study analyzed whether land (area) allocation decisions of cashew nut and sesame producers in southeast Tanzania have any bearing to changes in prices and non-price factors particularly in the long run. Variations in the annual production of these crops are attributed to the area cultivated and the yield per area of the crops. Production, yield and total area allocated to the crops are therefore of research importance.

19 5 As aforementioned, commodity prices play a crucial role in attaining efficient allocation of resources used in production, including land. The question is; how have commodity prices affected cashew and sesame acreage allocation over the years? In addition to price incentives, other non price factors like rainfall and technology are equally important in influencing agricultural production. The study used the Nerlovian adjustment model to cashew and sesame acreage in two districts of Southeast Tanzania (Nachingwea and Mtwara rural). Extension personnel and other development agents may use findings from this study to improve advisory services to farmers in terms of crop choice and crop mix, hence guide resource allocation. The findings are also expected to provide inputs into the planning of DADPS and in due course feed into improving planning for the ASDP when it is revised. 1.3 Objectives of the Study General objective The overall objective of the study was to assess farmers acreage responsiveness to price changes and to non-price factors for sesame and cashew in Southeastern Tanzania Specific objectives: i. To estimate growth in area, yield and production of sesame and cashew in Nachingwea and Mtwara rural over the period of 1995 to ii. To evaluate the impact of price on acreage of sesame and cashew in Nachingwea and Mtwara rural Districts. iii. To estimate the short and long-run elasticities of sesame and cashew in Nachingwea and Mtwara rural Districts.

20 6 1.4 Hypotheses In relation to the objectives as given above, the study worked on the following hypotheses: i. The null hypothesis in relation to specific objective number one was; There was no growth in area, yield and production of sesame and cashew in the study area during period. Mathematically; given the general exponential functional form Y = ae bt, then Ho: b = (1) Where: b = The coefficient for the growth rate ii. The second null hypothesis which corresponds to specific objective number two states that change in commodity prices did not influence farmers to adjust the area under sesame and cashew. Mathematically; Ho: d is d ic ) Where: d is = Adjustment coefficient for sesame in i th district, d ic = Adjustment coefficient for cashew nut in i th district. iii. The third null hypothesis specifies that the production of sesame and cashew in the study area was not price elastic, both in the short and long-run. Mathematically; Ho: sj = lj =0.... (3) Where: sj = Short run price elasticity of crop j, lj = Long run price elasticity for crop j.

21 7 These hypotheses were addressed using techniques discussed in chapter three. The results of the analyses were then summarized in tables, figures and in narrative form. In the next chapter a review of literature on different aspects of the study is presented.

22 8 CHAPTER TWO 2.0 LITERATURE REVIEW This chapter presents a review of the literature on the agricultural sector in Tanzania, examining policies governing the agricultural sector, and the production status of sesame and cashew nut sub-sectors in the country. A number of empirical studies on acreage response have been conducted both in developed and developing countries. The chapter also reviews some of these studies with a view to provide an understanding of models that are used in the analysis of acreage response. 2.1 Agriculture Sector in Tanzania The agricultural sector in developing countries plays a great role in terms of employment of the labour force and providing means of livelihood to a large proportion of the population (Nosheen and Iqbal, 2008; Nosheen et al., 2011; and Gurikar, 2007). In Tanzania, agriculture remains one of the sectors whose contribution to the national economy is relatively large (23.7% of the GDP), and thus, its productivity has a considerable impact on income levels of the farming communities. While the agricultural sector remains central to Tanzania s economy, its contribution to GDP has dropped from about 30% in 1998 to approximately 23.7% in 2008 (URT, 2009; URT, 2012). This could be the result of the demand of non-agricultural goods, as well as the post-farm gate economic activities such as value adding and petty businesses which are not taken into account of the agricultural GDP share (Tey et al., 2010). The decline of agricultural GDP implies that other sectors of the economy are growing. This is supported by Morrissey and Leyaro (2007) who argue that balanced growth is achieved if agriculture is increasingly commercialized while the manufacturing sector grows. In this sense, the

23 9 manufacturing sector and the economy will become diversified. However, Tanzania has not achieved this, hence, the economy continues to be agricultural-based. Despite Tanzania s agricultural potential in terms of availability of arable land and water sources for irrigation, agricultural growth has not been encouraging and its contribution to poverty reduction and industrial development has been low (Kidunda, 2010). For instance, in the period the sector s growth rate averaged 4.4% against MKUKUTA s target of sustained agricultural growth of 10% by 2010 which was not met (URT, 2009). Several constraints are responsible for such poor performance. These include: low net return due to high production costs, weak supporting institutions and poor infrastructure to mention a few. If the growth of the sector is to increase, and poverty and food insecurity are to be reduced especially in rural areas, then there should be deliberate efforts to increase agricultural productivity, coupled with efficient market mechanism. Realizing its importance, Tanzania has elected to foster agricultural growth through programmes like ASDP and Kilimo Kwanza, which strive to revive agriculture by attracting more investment to the sector as well as providing incentives to both small holder farmers and large investors. 2.2 Agricultural Policy in Tanzania In Tanzania agriculture is guided by the agricultural policy of 1997 which is in the process of being revised. The general objective of the policy is to improve the well being of the agrarian community whose livelihood is largely dependent on agriculture (URT, 1997). Since the majority of these people are smallholder farmers, their mode of production is not purely commercial though they also sell a small surplus. In light of this,

24 10 the agricultural policy of 1997 focused on commercializing agriculture to make it a profitable occupation, thereby increasing the income levels of smallholders. The policy addresses nine general objectives some of which are; (i) to improve the standard of living in rural areas through increased income generation from agricultural production, processing and marketing, and (ii) to increase foreign exchange earnings for the nation by encouraging the production and export of cash crops, food crops, as well as by-products and residues (URT, 1997). General economic reforms took place between 1986 and 1995 where the macroeconomic environment of Tanzania changed from a state controlled economy to a free market economy, which entailed changes of the guiding policies. While the macro-economic policy had a large positive effect on domestic commodity prices, the producer s share of the export price declined over this period for some key export crops, indicating that the agricultural sector policy did little to improve export price incentives during the late 1980s and early 1990s (URT, 2005). From the mid 1990s, significant changes took place for marketing institutions of major export crops (cashews, coffee, cotton, tea and tobacco). The changes included allowing private companies to enter into crop marketing. Primary societies now had a choice to sell to private traders or to cooperative unions. These changes resulted into producers receiving a higher share of the export price, which increased from an average of 54% during to 63% during , varying by crop (URT, 2005). However, during the late 1990s, these benefits to farmers were reduced following significant appreciation of the exchange rate (IMF, 2005) as presented in Fig. 1.

25 11 Figure 1: Trend of exchange rates of the Tanzanian shillings against the USD Source: Economic bulletins of Bank of Tanzania Sesame and cashew nut have been contributing to foreign exchange earnings for the country as well as to the livelihood of individual producers. Unshelled cashew nuts ranked fourth while sesame occupied the ninth position in list of major export crops of Tanzania in 2008 (Table 1). Table 1: Major export crops of Tanzania in 2008 Rank Quantity (Tonnes) Value (1000 $) Unit value ($/tonne) Commodity Tobacco, unmanufactured Coffee, green Cotton lint Cashew nuts, with shell Tea Wheat flour Dry peas Cotton carded, combed Sesame seed Palm oil Cashew nut shelled Cocoa beans Source: FAOSTAT, 2011

26 12 The two crops are particularly important as cash crops for farmers in southeastern Tanzania (Lindi and Mtwara Regions). In light of this, there is a need to regularly review the policy instruments which affect agriculture in general and sesame and cashew in particular, so as to improve the performance of the sector and enhance the livelihoods of farmers in the major growing areas. 2.3 Sesame Production Tanzania is one of the world s major sesame producers (FAO, 2005), ranking twelfth in the world and sixth in Africa after Sudan, Uganda, Nigeria, Ethiopia, and Central African Republic (Table 2). In Tanzania, Sesame (Sesamum indicum) is one of the common oilseed crops. It is the second most important cash crop in South Eastern Table 2: World s major sesame producers S/N Country Area Harvested Production Productivity (in "000" acres) (in "000" tonnes) (Tonnes/acre) 1 China India Myanmar Sudan Uganda Nigeria Pakistan Ethiopia Bangladesh Central African Republic Thailand Tanzania Egypt Guatemala Chad Paraguay Source: FAO, 2005

27 13 Tanzania after cashew (Mkamilo, 2004). Despite the release of several improved sesame varieties by Naliendele Agricultural Research Institute (NARI), sesame production in the zone is much lower than the potential (NARI, 2008). In order to realize this untapped potential, sesame production could be enhanced through use of improved technology as well as increased area under cultivation. The discussion on cashew sub sector is presented in section Cashew nut Sub-sector in Tanzania Cashew nut is grown in several regions of Tanzania including Mtwara, Lindi, Ruvuma, Coast, Dar es Salaam, and Tanga. Cashew nut is an important export crop for Tanzania (World Bank, 2004) and the biggest contributor to the economies of Mtwara, Lindi and the Coast Regions (Fynn, 2004), constituting a vital source of cash earning and a major means for improving livelihood of people in the respective areas. Despite its importance, the cashew nut sector nearly collapsed in the 1980s following replacement of primary societies, which acted on farmers behalf, by village agents acting for the Cashew nut Authority of Tanzania (CATA). This weakened farmers influence and brought the marketing of cashew under the control of the government (Baregu and Hoogeveen, 2009). The start of villagization in the mid 1970s also contributed to the decline in cashew production as farmers abandoned their cashew fields. However, the crop has since then made a remarkable recovery (Fig. 2) following economic reforms which begun in 1986 and sector reforms which took off in the mid-1990s (World Bank, 2004). Economic reforms included trade liberalization and exchange rate adjustment, while sector reforms included eliminating the monopoly of the Tanzania Cashew nut Marketing Board (TCMB).

28 14 Figure 2: Production of cashew nut in Tanzania: Source: FAO stat (cited 18/07/2012). The regulatory functions of the TCMB were taken over by the Cashew nut Board of Tanzania (CBT) (Mkude, 2003). Exchange rate adjustment resulted into tripling the real producer prices for raw cashew nut from the mid-1980s to the mid-1990s (World Bank, 2004). These reforms were designed to improve producer prices. It was necessary however to empirically assess how farmers responded to price through area allocation and subsequent productivity and production growth. The next section presents the discussion on growth rates and the concept of acreage response in agriculture. 2.5 Growth Rate Analysis and Acreage Response Compound growth rates in agriculture The concept of computation of compound growth rates has been used quite widely, particularly in the discipline of agricultural economics (Gurikar, 2007; Ramachandra, 2006; and Guledgudda, 2005). The current general procedure being followed for

29 15 computation of growth rate is presented in this section as adapted from Prajneshu and Chandran (2005). If Y t denotes an observation (eg. Agricultural production, yield, or area) at time t and r is the compound growth rate, estimating the growth rate is based on the model presented below: Y t = Y 0 (1 + r) t e (4) The model is usually linearlized by logarithmic transformation, resulting into equation (8). Ln (Y t ) = Ln (Y 0 ) + Ln (1 + r)t (5) The data are then used to fit equation (8) using ordinary least square (OLS). The model s goodness of fit is assessed by the coefficient of determination (R 2 ), as well as the sign and levels of significance of respective coefficients. The compound growth rate is given by the antilog of log (B-1)100 where B = Ln (1 + r). Many studies have estimated the trends for area, yield and production of different crops using different forms of growth models. The exponential model is the most commonly used. Ramachandra (2006) evaluated the trend in area, production and productivity of sapota fruit in Karnaka state, India using a growth model of the form; Y t = ab t t.(6)

30 16 Where; Y t = area or production or productivity in the year t a = intercept indicating Y in the base period (t = 0) b = Regression coefficient t = Time period in years t = Disturbance term for the year t. The linearlized form of equation 6 was used for empirical estimation (Equation 7). lny t = ln a + tln b + ln t.. (7) Where Y t, a, b, t and U t are as previously defined. The findings by Ramachandra showed a positive growth rate of 4.54% per annum while production and productivity showed negative growth rates of -1.98% and -6.24% per year, respectively. Gurikar (2007) similarly used a growth model in the linearlized form of equation (8) as presented by equation (9) to study rates of change in the area, production and productivity of onions in major growing states of India. Y = ab t e.... (8) Where; Y= dependent variable (Area or production or productivity) a = intercept term b = (1+r) and r is the compound growth rate.

31 17 t = time trend (t = 1,2,3,..n) e = Natural log error term. log Y = log a + t log b +....(9) Where; all variables are as previously defined. Results showed that Karnataka state registered a significant positive growth rate in the production of onion, resulting from increased area in the major onion growing districts. Likewise, Khan et al. (2002) used an exponential growth model in equation (10) in their analysis of growth and trend of production and yield of two varieties of rice in Bangladesh. Y = ae bt..(10) Where; Y = dependent variable (production or yield) t = independent variable (time) a = intercept b = growth rate. The final fitted model took the logarithm form of: ln Y = lna + bt...(11) with b representing the growth rate in decimals which was multiplied by 100 to express the annual percentage compound growth rate. The findings by Khan et al. (2002) indicated that the two rice varieties registered positive growth rates both in production

32 18 and yield. Studying the acreage response to price and non-price factors is equally important, and it is discussed in the next section Acreage response and estimation techniques Acreage response to commodity price is among the core issues of agricultural economics (McDonald and Sumner, 2002). Assessing acreage response is important in analyzing farm programs. Many studies have been done to assess the relationship between crop production (acreage, productivity and production) and crop prices. According to Houston et al. (1999), acreage has been measured in different forms including; the planted acreage, harvested acreage, volume or weight produced of an individual crop, and revenue per planted unit. The planted acreage of any crop may increase or decrease in response to changes in the price per unit the crop is expected to bring (Barnes, 2010). This relationship is the primary principle behind acreage or supply response. Many studies have been conducted on supply and/or acreage response to commodity price and other non-price factors. The literature shows that Nerlovian adjustment also known as the expectation model is the most widely used estimation technique by researchers. For example, Tey et al. (2010) used a logarithmic functional form of the linear Nerlovian expectation model to study the acreage response of paddy in Malaysia. The area planted with paddy was used as the dependent variable while lagged price, lagged area, lagged yield and government support were independent variables in the model. The negative short run elasticities of area under paddy with respect to paddy price was and that for yield was indicating that there was an inverse relationship between area under paddy and paddy price. Similarly, the long run price elasticity of indicated a negative relationship between the planted area for paddy and price. The study also found that the expectation coefficient of paddy price was one, meaning that the expected price

33 19 for the current year was same as the actual price in the previous year. Likewise, the area adjustment coefficient was which meant, farmers adjusted moderately toward the desired area planted with paddy. Another study by Nosheen and Iqbal (2008) in Pakistan also applied a Nerlovian model to estimate the response of cotton, wheat and sugarcane area to changes in their respective prices and other relevant factors. The coefficients of the area response were estimated using Ordinary Least Squares (OLS). The short run price elasticity of area for cotton was estimated to be while the long run price elasticity was These results suggest that the price of cotton played an important role in farmers decisions to expand or contract the area for cotton during the reference period. The adjustment coefficient was found to be indicating farmers slow pace of adjustment in response to price movement in the long run. Meanwhile, the short run price elasticities of wheat and sugarcane area were and 0.229, respectively while the long run elasticities were calculated at and 0.653, respectively implying that sugarcane was relatively more elastic than wheat. One of the reasons for relatively low elasticity coefficients for wheat may be the very large area already devoted to wheat cultivation in Pakistan and its dominance in the cropping pattern and thus not leaving much scope for further extension of wheat area (Nosheen and Iqbal, 2008). These results point out the influence of producer prices on the area planted to wheat and sugarcane. The findings also revealed that the area adjustment coefficients for wheat and sugarcane were 0.44 and 0.35, respectively, meaning that farmers could adjust more easily to wheat area than they could for sugarcane. The possible explanation for this

34 20 relatively slow adjustment in sugarcane crop may be its longer duration to maturity and ratooning practices. These findings could have similar implication for sesame and cashew under study. One would expect that the area adjustment to commodity price changes would be faster for sesame relative to cashew, thus a higher adjustment coefficient. Other studies have added innovation in the analytical models. Duffy et al. (1994) for example used an expected utility model that included output price and yield uncertainty to estimate cotton, maize and soybean acreage response for South-Eastern USA. Time series data were used and the model fitted the soybean and maize data well. Own-price elasticity estimates were and for maize and soybeans, respectively, which were higher than the national average, implying the availability of production substitutes in the southeastern part of USA. This implies producers were responsive to changes in profitability, following product price changes. Mythili (2008), while studying acreage and yield response for major crops in the pre- and post-reform periods in India, included the lagged relative price index and rainfall as quantitative variables in the final functional form which was based on the Nerlovian model. A crop dummy variable was added to capture the effect of the main and substitute crops, and a period dummy to reflect the period before and after the reforms. McKay et al. (2006) studied aggregate export and food crop supply response in Tanzania using Nerlove s model by incorporating price expectations and adjustment costs. Their estimates suggested that agricultural supply response was quite high indicating that the potential for agricultural sector response to liberalization of agricultural prices and marketing was quite significant. Generally speaking, variants of the Nerlovian model have been used in different forms by researchers to address supply or acreage response of various crops given changes in both price and non-price factors (Rahji et al., 2008;

35 21 Niamatullah and Zaman, 2009; Molua, 2010; and Nosheen et al., 2011). These and other researchers have used different variables in their models and they are discussed in the next section Variables used in acreage response studies Dependent variables Farmers tend to plan certain levels of output to be produced in response to price and nonprice factors. The planned output represents farmer s response to price expectations. It is, however, difficult to find time series data on planned output, hence the need to use an appropriate proxy. There is an ongoing debate among researchers about which dependent variable to use. One group of researchers including Mythili (2008) contends that the area under a crop could be an appropriate proxy for planned output on the basis that data on area under different crops can be readily obtained from various sources. Other researchers claim that production response (total output) is the most appropriate variable to measure farmers response to price and other policy induced changes rather than the area of a particular crop. They stress that in modern agriculture with land saving technologies, land is regarded as a secondary factor in production (Gurikar, 2007). Significantly higher values of production can be obtained from a small area if intensive technologies such as green houses and liquid or foliar fertilizers are used. Yet, there is another group of researchers who suggested using both area and yield in addressing farmers response to price and non-price factors (Ibid). In the present study, area under the crop was used as a dependent variable to assess the farmers response to price and non-price factors since land is the primary factor in production of cashew nut and sesame in the study area. Land saving technologies which

36 22 include green house and foliar fertilizers are not being used for the production of sesame and cashew in the study area. In addition, crop area has been preferred over production as farm production is also influenced by weather conditions, which are often beyond the control of farmers. Yield was not used because it is also subject to more random variation than acreage due to factors outside the farmers control such as the weather (Nosheen and Iqbal, 2008). Based on the literature and discussion with farmers the main factors (independent variables) which impact farmer s allocation of crop area are presented in the next section Independent variables Independent variables which are commonly used in acreage response studies include price and non-price factors. Of these explanatory variables, commodity prices have frequently been used for reasons discussed below. According to Minde (1991); and Rweyemamu and Kimaro (2006), producer prices are among the most important and effective tool for influencing agricultural output. These prices are crucial in determining profit or loss in the farm enterprise. When producer prices are calculated in relation to the costs incurred by farmers in the production process, they lead to profit and provide incentive to producers to grow more (Gurikar, 2007). This has led to increased attention on the effect of short-run changes in prices on production behaviour. Gurikar further asserts that in order to bring about sustained and balanced economic growth, it is very important to understand the long-run effect of prices on production. Different forms of price factors have been used to study farmers supply responses. For instance, Nosheen et al. (2011) used prices of a commodity received in the recent past to study farmers response to price and other factors for rice production in Pakistan. The results showed positive price elasticity of acreage. In Cameroon, Molua (2010) used

37 23 relative prices in his study on price and non-price determinants of acreage response on rice. He concluded that the area under rice would increase by 1.35% for a ten percent increase in world price relative to rice producer price. Meanwhile, Niamatula and Zaman (2009) used lagged market price to estimate the Nerlovian adjustment model in studying the acreage response of wheat and cotton to respective price changes. Their results revealed that short and long run price elasticities of wheat production were and , respectively. Lagged prices of a commodity are often used in the model because prices received by the farmers in the recent past shapes economic incentive for the commodity. This is supported by Nosheen and Iqbal (2008) and Nosheen et al. (2011) who argue that farmers resources allocation decisions are mainly based on the crop prices they received in the recent past. However, price differences explain only part of the variation in the response variable. Acreage response has also been considered a consequence of changes in several non-price factors, which influence production. For this reason a favourable price policy alone may not influence farmers to increase agricultural output through increased acreage. For instance, it is known that yield is an important determinant for the profitability of crops in a given year. Yet the yield of any crop at its planting time is not known. Farmers therefore base their expectation of profitability of a given crop on the yield realized in the recent past. Hence lagged yield enters the model as an independent variable. Lagged area is also used as an independent variable to capture the effects of farmers know-how and experience with the given crop (Molua, 2010; Nosheen and Iqbal, 2008; and Nosheen et al., 2011). Other non-price factors are known to influence agricultural production. According to Gurikar (2007), changes of climatic factors as well as incidence of pests and diseases

38 24 adversely affect agricultural production in the short-run while technological advancements cause long-run supply changes. Rainfall has been used as an independent variable quite frequently in empirical studies to represent non-price factors (Gurikar, 2007; Mythil, 2008; Niamatula and Zaman, 2009; Molua, 2010). Meanwhile, technology has been represented by trend variable (Gurikar, 2007) to reflect its tendency to change over time. Different analytical innovations have been developed to accommodate the other non-price factors. Based on the experience of other researchers as discussed in this chapter the methodology and analytical model for this study is presented in chapter three.

39 25 CHAPTER THREE 3.0 METHODOLOGY 3.1 Conceptual Framework This conceptual framework focuses on the analysis of cashew nut and sesame supply response relations. Basically, supply response concerns output response to a change in price of the product (Akanni and Okeowo, 2011). This may be due to the application of more or less productive resources resulting from a price increase or decrease. The response may also be represented by a change in farm size. Variables like credit, price, weather, and market information which influence agricultural production are expected to bring about supply response. Under peasant economy, farmers tend to practice intercropping or cultivate a number of plots in different locations to diversify income and reduce risks of crop failure (Mkamilo, 2004). The underlying objective for pursuing these objectives is utility maximization. It is known that farmers respond to price incentives partly through intensive application of variable inputs to a given piece of land (Bridges and Tenkorang, 2009). It is also well established that agricultural policies and institutional factors influencing commodity prices and input prices have an impact on farmers decisions on resource allocation. This also includes assigning acres of land to different crops. Since the product price is the most important variable in the response function, policy makers should take into account farmers response to price during policy making. 3.2 Research Design In Tanzania, most of the agricultural policies are formulated at the national level, and implemented through various programmes such as the ASDP, Kilimo Kwanza, which

40 26 conform to regional and global programmes such as the Comprehensive Africa Agriculture Development Programme (CAADP). Within each of these programmes, there is room for LGAs to adapt the macro policies and strategies to suit their specific requirements. Adaptations can be made by LGAs to provide room for developing policies and strategies within the framework of the District Development Programme and DADPS. These adaptations include; (i) efforts to improve farmers access to and use of agricultural knowledge, technologies, marketing systems and infrastructure; (ii) promoting stronger links between smallholders and agribusiness; and (iii) raising the amount and quality of cash and food crops produced to improve the livelihood of the farming communities. This study used district level time series data for acreage, production and prices of cashew nut and sesame to assess the farmers response to changes in policies governing commodity prices as well as other non-price factors. 3.3 Choice of Crop and the Study Area Nachingwea and Mtwara rural Districts are found in the southeast of Tanzania. Nachingwea District is in Lindi Region while Mtwara rural is in Mtwara Region. These districts were selected for this study due to their climatic and topographical characteristics which suits the production of cashew nuts and sesame. It is not surprising to find farmers in these districts cultivating both crops because the soils and climate are suitable. Cashew nut and sesame were selected because they constitute the most important cash crops and they play a substantial role in generating income for the majority of the community members in the zone (Fynn, 2004).

41 Description of the Study Area Area and location The Directorate of Research and Development of the Ministry of Agriculture, Food and Cooperatives has divided Tanzania in seven research zones including Southeastern zone. The zone which is in the south east of Tanzania comprises of Mtwara and Lindi Regions; and Tunduru District which is in Ruvuma Region. It covers sq km of which sq km is in Mtwara Region and sq km is in Lindi Region; while Tunduru District covers the remaining sq km. The zone borders with the Indian Ocean in the east, Morogoro and Ruvuma Regions in the west, the Coastal Region in the North and with the Ruvuma River in the south, which forms a boundary with Mozambique (Fig. 3). Mtwara rural District covers 3597 square kilometres. The Indian Ocean borders it to the East, while the Ruvuma River separates it from Mozambique to the South. The district borders with Lindi Region to the North and Tandahimba District to the West. Meanwhile, Nachingwea District which is one of six districts of Lindi Region is bordered to the North by Liwale District, and to the east by Ruangwa District. To the Southeast, it borders with Mtwara Region, while Ruvuma Region shares the border with it to the Southwest Demography Based on 2012 Tanzania National Census, the size of the Mtwara District population is of which are males and are females (URT, 2013). The Census results also reported the total population of people in Nachingwea District with males and females. Mtwara Region recorded a higher average annual intercensal population growth rate, censuses (1.2%) compared to Lindi Region (0.9%) in which Nachingwea District belongs.

42 Climate and topography Generally, south-eastern Tanzania is characterized by mixed farming systems whose elements change with variations in climate and environment. There are two main seasons: a humid and hot wet season (November to May) and a cooler, less humid dry season (June to October). Figure 3: A map of Southeastern Tanzania showing sample districts (Nachingwea and Mtwara rural) The mean annual rainfall ranges from 800 mm within inland and central areas to 1200 mm in the hills and plateau near the Coast. Soils are variable, ranging from deep, well drained, but not very fertile sandy soils of the sedimentary zones to deep, well drained, and somewhat more fertile red clay soils of Nachingwea and Masasi Districts